Which catchment characteristics control the magnitude-frequency distribution of river discharge?

The incision of rivers into bedrock is one of the most important processes driving long-term landscape evolution. River incision is non-linearly dependent on river discharge, because discharge must exceed a certain threshold to be able to transport bedload sediment and expose the riverbed to erosion. Fluvial erosion efficiency is, therefore, strongly controlled by the magnitude-frequency distribution of river discharge. Previous studies have found non-uniform relationships between precipitation and runoff, demonstrating that catchments strongly modulate the discharge response after a precipitation event. This has implications for how rainfall drives fluvial erosion.

We present initial results from a study focusing on a climate gradient in Chile to investigate which catchment characteristics control the magnitude-frequency distribution of river discharge. We exploit a large dataset of 809 daily discharge records covering Chile from 18°S to 55°S, gauged by the 'Dirección de General de Aguas'. Based on extensive quality checks, we removed stations downstream of lakes and discharge records with <10 years of data or large data gaps, yielding a total of 400 suitable records.

We used the weighted sum of 1 to 3 inverse gamma functions fitted to the data to represent discharge variability of different flow regimes. Results show higher discharge variability in arid and alpine catchments, as compared to humid catchments. In most catchments (99%), the data was best described by 2 or 3 weighted inverse gamma functions, revealing that most basins have multiple discharge regimes. Next, we explore the relationship between the distribution of discharge and catchment characteristics such as climate, vegetation density, regolith thickness, and lithology, using a streamflow recession model. Drawing on recent theories that link the distribution of discharge to fluvial incision rates, we will explore how these catchment characteristics may influence long-term fluvial incision rates in unexpected ways.